Skip to content

Data Dictionary

Every census tract profile includes 37 fields across five categories. This page provides clinical thresholds and interpretation guidance for each field.

Quick triage

If you only have time to check one number, look at sdoh_index. Values above 0.6 indicate high social vulnerability — patients from these tracts are likely to face multiple compounding barriers to health.


Demographics (American Community Survey)

Source: US Census Bureau, American Community Survey (ACS). Updated annually.

Field Type Unit Typical Range Clinical Threshold Interpretation
total_population int persons 500–15,000 <1,000 = unreliable rates Denominator for all rate metrics. Small populations produce less reliable estimates.
median_household_income float dollars $15,000–$150,000 <$30,000 = high risk Predicts healthcare access and outcomes. Low incomes correlate with chronic disease, delayed care-seeking, and medication non-adherence.
poverty_rate float % 0–60% >20% = high-poverty Strongest single predictor of poor health outcomes. Associated with increased chronic disease, limited healthcare access, and reduced life expectancy. Screen for food insecurity, medication affordability, and transportation barriers.
uninsured_rate float % 0–35% >15% = access barriers Directly impacts healthcare utilization and preventive care. Uninsured individuals are more likely to delay care, skip medications, and present with advanced disease.
unemployment_rate float % 0–25% >10% = economic distress Associated with depression, anxiety, substance use, and loss of employer-sponsored insurance. Consider behavioral health screening.
median_age float years 18–65 >45 = chronic disease risk Tracts above 45 may have higher chronic disease prevalence. Tracts below 25 may indicate student populations or young families with pediatric needs.

Vulnerability (CDC/ATSDR Social Vulnerability Index)

Source: CDC/ATSDR Social Vulnerability Index (SVI). All values are national percentiles (0–1). Higher = more vulnerable.

Clinical rule of thumb

SVI percentile above 0.75 = top 25% most vulnerable nationally. These communities need intensive social needs screening and care coordination.

Field What It Measures Interpretation
svi_themes.rpl_theme1 Socioeconomic status: poverty, unemployment, no insurance, no high school diploma, housing cost burden Percentiles above 0.75 indicate compounding barriers — poverty, low education, and lack of insurance — that drive worse outcomes across virtually all conditions.
svi_themes.rpl_theme2 Household composition & disability: aged 65+, under 17, civilian with disability, single-parent households High percentiles indicate populations with greater care dependency. May need home health services, accessible facilities, and caregiver support.
svi_themes.rpl_theme3 Minority status & language: racial/ethnic minorities, limited English proficiency Consider culturally competent care, interpreter services, and awareness of health disparities affecting specific racial/ethnic groups.
svi_themes.rpl_theme4 Housing type & transportation: multi-unit/mobile homes, crowding, no vehicle, group quarters Patients may miss appointments due to lack of transportation, live in crowded conditions promoting infectious disease, or face housing-related health hazards.
svi_themes.rpl_themes Overall composite across all 4 themes The single best summary of community vulnerability. Above 0.75 = most vulnerable 25% nationally. Use as a quick triage metric.

Health Outcomes (CDC PLACES)

Source: CDC PLACES. All values are crude prevalence percentages among adults 18+, derived from Behavioral Risk Factor Surveillance System (BRFSS) model-based estimates.

Field What It Measures Unit Typical Range High-Burden Threshold
places_measures.diabetes Diagnosed diabetes % 5–25% >12%
places_measures.obesity BMI >= 30 % 15–50% >35%
places_measures.mhlth 14+ days mental distress/month % 8–25% >16%
places_measures.phlth 14+ days physical distress/month % 5–20% >15%
places_measures.bphigh Hypertension (high blood pressure) % 20–50% >35%
places_measures.casthma Current asthma % 5–15% >10%
places_measures.chd Coronary heart disease % 2–12% >7%
places_measures.csmoking Current smoking % 8–30% >20%
places_measures.access2 No health insurance (18–64) % 3–30% >15%
places_measures.checkup Annual checkup % 55–85% <65% (low = concern)
places_measures.dental Annual dental visit % 35–80% <55% (low = concern)
places_measures.sleep Short sleep (<7 hours) % 25–50% >38%
places_measures.lpa No leisure-time physical activity % 15–45% >30%
places_measures.binge Binge drinking % 10–30% >20%

Clinical context for key measures

Diabetes (places_measures.diabetes)

Prevalence above 12% indicates a high-burden area. Diabetes drives cardiovascular disease, kidney disease, and amputations. High-prevalence tracts may benefit from community-based diabetes prevention programs, A1c screening, and nutrition counseling.

Mental health distress (places_measures.mhlth)

Prevalence above 16% signals significant community mental health burden. Correlates with substance use, suicide risk, and reduced workforce participation. Screen for depression and anxiety; assess behavioral health service availability.

Hypertension (places_measures.bphigh)

Hypertension is the leading modifiable risk factor for cardiovascular disease and stroke. Prevalence above 35% warrants community-level blood pressure screening programs and medication adherence support.

Smoking (places_measures.csmoking)

Smoking above 20% indicates a high-burden area. Leading preventable cause of death — drives lung cancer, COPD, and cardiovascular disease. Prioritize tobacco cessation programs.

Checkup and dental rates

These are inverted — low values are concerning. Checkup rates below 65% and dental rates below 55% suggest underutilization of preventive care, often indicating access barriers, distrust of healthcare, or competing priorities.


Environmental (EPA EJScreen)

Source: EPA EJScreen. Environmental justice screening indicators at the census tract level. Data may come from the EPA API (_source: "ejscreen_api") or be estimated from demographic correlations (_source: "estimated").

Field What It Measures Unit Interpretation
epa_data.pm25 Fine particulate matter (PM2.5) μg/m³ Annual average concentration. EPA standard is 12 μg/m³; above indicates poor air quality.
epa_data.ozone Ground-level ozone ppb 8-hour average. EPA standard is 70 ppb. Triggers respiratory issues.
epa_data.diesel_pm Diesel particulate matter μg/m³ Proxy for traffic-related air pollution. Higher values near highways and industrial areas.
epa_data.air_toxics_cancer_risk Lifetime cancer risk from air toxics per million EPA considers >100 per million elevated.
epa_data.respiratory_hazard_index Respiratory hazard from air toxics ratio Above 1.0 indicates potential respiratory concern.
epa_data.traffic_proximity Traffic proximity and volume vehicles×distance Higher values = more traffic exposure.
epa_data.lead_paint_pct Pre-1960 housing proportion 0–1 Indicator of lead paint risk. Above 0.5 = significant concern.
epa_data.superfund_proximity Proximity to Superfund sites count/distance Higher = closer to contaminated sites.
epa_data.rmp_proximity Proximity to RMP facilities count/distance Risk Management Plan facilities with hazardous chemicals.
epa_data.hazardous_waste_proximity Proximity to TSDF facilities count/distance Treatment, storage, and disposal facilities.
epa_data.wastewater_discharge Wastewater discharge indicator toxicity-weighted Stream proximity weighted by pollutant toxicity.

Data source indicator

The _source field in epa_data indicates whether values come from the EPA EJScreen API ("ejscreen_api") or are estimated from demographic correlations ("estimated"). Estimated data provides reasonable approximations but should be interpreted with caution.


Composite Index

Field Type Scale Interpretation
sdoh_index float 0–1 Computed from normalized poverty rate, uninsured rate, unemployment rate, and SVI overall percentile. Above 0.6 = high vulnerability. The single most useful triage metric for clinical risk assessment. Use to prioritize social needs screening, care management enrollment, and community health worker referrals.

Data Sources & Methodology

Source Coverage Update Frequency Geographic Level
American Community Survey (ACS) All US census tracts Annual (5-year estimates) Census tract
CDC/ATSDR SVI All US census tracts Biennial Census tract
CDC PLACES All US census tracts Annual Census tract

Census tract = a statistical subdivision averaging ~4,000 people, designed to be relatively homogeneous in population characteristics, economic status, and living conditions. This is the standard geographic unit for neighborhood-level health analysis.

SDOH index computation: The composite sdoh_index normalizes poverty rate, uninsured rate, and unemployment rate to 0–1 scales, then averages them with the SVI overall percentile (rpl_themes). The result is a single 0–1 score where higher values indicate greater social vulnerability.


Programmatic Access

All field definitions are available programmatically via the /v1/dictionary endpoint, which returns structured JSON with type, source, clinical relevance, and example values for every field.